What Is a Battery Protection Circuit?

Detailed Explanation

A lithium-ion cell without protection is a genuine safety hazard. Unlike alkaline or NiMH batteries, Li-ion cells can reach thermal runaway — an uncontrollable exothermic reaction that produces heat, smoke, fire, and toxic gas — if they are overcharged, deeply discharged and then recharged, or short-circuited at high current. A protection circuit is the primary hardware safeguard against all three failure modes.

The Standard Single-Cell Protection Circuit

The most common implementation for a single Li-ion cell is a protection IC paired with two back-to-back N-channel MOSFETs:

                    [Cell +] ──── [Protection IC] ──── [MOSFET D (discharge FET)] ──── [MOSFET C (charge FET)] ──── [Pack +]
                    [Cell −] ──────────────────────────────────────────────────────────────────────────────────────── [Pack −]

The two MOSFETs are placed in series in the current path, with their body diodes oriented in opposite directions:

• Discharge FET (D): body diode points from source to drain — allows current to flow into the cell from a charger when this FET is off, even if the discharge path is cut.
• Charge FET (C): body diode points the other way — allows current to flow out to the load even when the charge path is cut (cell near empty and being discharged below threshold).

This back-to-back arrangement means the protection IC can independently cut the charge path (stop charging when overcharge is detected) without blocking the discharge path, and vice versa.

Fault Conditions and Response

Protection IC examples and characteristics:

The DW01A + FS8205A Implementation

The DW01A + FS8205A pairing is the most commonly seen protection circuit in single-cell consumer devices:

• DW01A: a tiny SOT-23-6 protection IC with internal reference, comparators, and delay timers
• FS8205A: a dual N-channel MOSFET in a single SOT-23-6 package (the two back-to-back FETs)

The implementation requires only a handful of passive components:

• The FS8205A's gate drive is controlled directly by the DW01A's OD (overdischarge/charge FET gate) and OC (overcharge/discharge FET gate) outputs
• A single resistor across the current-sense pin sets the short-circuit threshold
• Decoupling capacitors on the DW01A's supply pins

Total BOM for this protection circuit: two ICs, approximately four passive components, assembled in a few mm² of board area.

What a Protection Circuit Does NOT Do

A single-cell protection circuit covers the basic safety thresholds. It does not:

• Balance multiple cells in series: if two cells in a 2S pack charge at different rates, one cell can reach 4.2V while the other is still at 3.8V. The protection circuit for each cell would individually protect it, but there's no mechanism to equalise them. A BMS handles this.
• Estimate state of charge (fuel gauging): the protection IC knows the cell voltage but not the remaining capacity. A fuel gauge IC (e.g. BQ27220, MAX17048) is needed to report remaining runtime or percentage to firmware.
• Control the charge process: the protection circuit passively watches and disconnects; the charger IC actively controls the CC/CV charging profile. Both are needed.

Design Considerations

• FET Rds(on) and current rating: the MOSFETs in the current path must handle the maximum continuous discharge current with acceptable voltage drop and power dissipation. An FS8205A can handle 6A typically; higher-current designs need larger FETs or parallel FET topologies. Always check the FET's Rds(on) at the actual gate drive voltage the protection IC provides — usually 4–5V, not 10V — which is typically higher (worse) than the datasheet's RDS on at 10V.
• Self-discharge through the protection circuit: the protection IC itself draws a small quiescent current (typically 3–10 µA). In a low-power product that sits unused for months, this slowly depletes the cell. Verify that the IC's quiescent current is acceptable for your product's storage lifetime.
• Short circuit response time: the delay between a short circuit event and the FET turning off is not zero. During this delay, the cell sees the full short-circuit current, which can be hundreds of amps for a few milliseconds. This is within spec for the cell and protection circuit in a well-designed implementation, but a fuse placed in series with the cell is an additional safeguard against complete MOSFET failure during a sustained fault.
• Placement near the cell: the protection circuit should be as close as possible to the cell's terminals. Long wires or traces between the cell and the FETs create additional resistance and inductance — the latter slows the protection IC's ability to detect current rise accurately.
• Production battery circuit design: integrating protection, charging, power path, and fuel gauging into a complete, certified battery management subsystem is a hardware design scope Zeus Design's engineering team handles routinely for wearable, industrial, and IoT products.

Common Mistakes

• Buying "protected" cells and assuming no additional protection is needed in the design: a protected cell has a protection circuit built into the cell itself, but that circuit is a minimum safety floor — it is not a substitute for a system-level protection circuit when the cell is integrated into a custom PCB design.
• Using a protection IC's maximum rated current as the design continuous current: the FETs' Rds(on) causes real heat dissipation at high current. A circuit rated at 3A maximum may run safely at 3A only with sufficient copper area and thermal management. Derate to 70% of maximum for continuous operation.
• Omitting protection on "just a prototype": prototypes are the devices most likely to be left on a bench charging unattended, shorted by a crocodile clip, or discharged to zero. Including the protection circuit from the first prototype costs a few dollars and prevents a far more expensive PCB fire or ruined equipment.
• Confusing the DW01's OD and OC pin functions: the labels are counterintuitive. OD ("overdischarge") drives the discharge FET gate; OC ("overcharge") drives the charge FET gate. Swapping these in a schematic produces a circuit that enables the discharge path on an overcharge fault and vice versa — exactly the wrong behaviour.